Hi Florian, Matthew,

> > > Here is a little something on the topic of thru-zero FM.
> > > http://www.uni-bonn.de/~uzs159/vco6.html
>
> > This looks like the Curtis CEM 3340 is a thru-zero VCO too. Am I
> > wrong or right

Unfortunately it is not. Oscillators like the CEM3340 and the A-110
require a 'standing current', positive for the 3340, negative for the
A-110. For normal 1V/octave (exponential) use this is constant, and
passes through an exponential converter (controlled by the freq CV)
and thence charges a capacitor. The capacitor is discharged suddenly
in the A-110, hence it's a sawtooth core; in the CEM3340, current
mirrors reverse the current and discharge the cap at the same rate as
it was charged, hence a triangle core. For linear FM control of both
of these, the linear FM input simply adds or subtracts current from
the standing current, by an amount linearly proportional to the
applied voltage: if this voltage becomes of such a size so as to
reduce the (magnitude of the) standing current to zero, the
oscillator simply stops, as there is nothing to charge the cap! To
prevent this happening generally requires quite a lot of trickery to
reverse everything at the appropriate moment, hence all the switches
and comparator in the Schmitz design.

Yet another thing on my list to look at once I finish my studies and
have the time!

Tim